Mailing-List: contact dev-help@cocoon.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@cocoon.apache.org
Message-ID: <3F4B6BA2.1070803@leverageweb.com>
Date: Tue, 26 Aug 2003 10:16:02 -0400
From: Geoff Howard <cocoon@leverageweb.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
 rv:1.3.1) Gecko/20030425
MIME-Version: 1.0
To: dev@cocoon.apache.org
Subject: Re: cvs commit: cocoon-2.1/src/blocks/html/lib .cvsignore
 jtidy-04aug2000r7-dev.jar
References: <LKENJPKDAIKAPKHEOCMMAEHKENAA.cziegeler@s-und-n.de>
 <03c801c36bce$abb02a70$1600a8c0@hypernexinc.local>
In-Reply-To: <03c801c36bce$abb02a70$1600a8c0@hypernexinc.local>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit

Roger I Martin PhD wrote:
> Hi Carsten,
> 
> Just a thought about something I've run into.  When a third party makes a
> block or webapp that depends on the same lib but a different version, can
> the jar indexing capability of jdk1.4 jar utility be employed to resolve the
> issues?  It involves placing and maintaining a correct Class-Path: ... in
> the jar's manifest and then running jar -i *.jar.
> http://java.sun.com/j2se/1.4.2/docs/tooldocs/windows/jar.html#i
> 
> Also does anyone know if the indexing really does or would affect the speed
> of bringing up a wepapp and say Tomcat?
> 
> Roger

I have no experience with this but have been interested in it.  Would 
this necessarily introduce a dependency on 1.4? (in which case I doubt 
we'd decide to go that way?)  Or would it only require jar-ing with 1.4
and writing backwards compatible classloaders which respect the indexing?

Geoff (reading the link now)